about color bias problem

[fm123a]

hi，
the color bias is related with the gain K ?, According to some experiments，As K gets bigger, that color deviation gets bigger。 And for camera params in this demo， as K gets bigger value, color bias only sample some values in given color bias,that is right? how to sample color bias for A bigger K ?

[Vandermode]

Since we record empirical color bias for different K, you may simply reorganize and sample them considering different K

[fm123a]

你好，还有个问题，看您论文，颜色偏差是随机采样的。因为我发现如果从gain值很小到gain值很大估计，对于这个颜色偏差，当gain值在一定范围（比如100到256） 是稳定的。当gain值越大，这个颜色偏差就变得非常大。那是否要在最小和最大gain值估计得到的颜色偏差范围去随机采样。我看你代码中相机参数，当增益K变化时，颜色偏差是一定范围的取值。

…

---Original--- From: "Kaixuan ***@***.***> Date: Thu, Jun 9, 2022 15:08 PM To: ***@***.***>; Cc: ***@***.******@***.***>; Subject: Re: [Vandermode/ELD] about color bias problem (Issue #26) Since we record empirical color bias for different K, you may simply reorganize and sample them considering different K — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: ***@***.***>

[Vandermode]

When you sample a small (large) K, you then sample color bias from empirical distribution formed by those recorded under small (large) K values.

[fm123a]

Thank you for your reply. I have another question about noise image synthesis.If the clean image is divided by the low light factor, then add noise, and then multiply by the low light factor。This operation results in synthetic noise amplification，and it does not accord with the real noise distribution? for example ,the low value is 60 ,The corresponding noise value is 20， if the low light factor is 100, After processing， the low value is (60/100 +20)*100=206, but original value is 60+20= 80 , it is right?

[happycaoyue]

Thank you for your reply. I have another question about noise image synthesis.If the clean image is divided by the low light factor, then add noise, and then multiply by the low light factor。This operation results in synthetic noise amplification，and it does not accord with the real noise distribution? for example ,the low value is 60 ,The corresponding noise value is 20， if the low light factor is 100, After processing， the low value is (60/100 +20)*100=206, but original value is 60+20= 80 , it is right?

wrong. the noise value depend on low value.